Monad of no `return` Proposal (MRP): Moving `return` out of `Monad`

[Henning Thielemann]

On Fri, 2 Oct 2015, Manuel Gómez wrote:

> In a post-AMP Haskell, however, retaining `return` in `Monad` does not 
> serve any purpose: any `Monad` instance requires an `Applicative` 
> instance, wherein `pure` must be defined, and it is a law required by 
> all instances that they behave identically.  The only expressive purpose 
> of `return` being in the `Monad` type class is providing an opportunity 
> to break the laws the community expects for all instances of `Monad`.

We will get constant breakage this way. You mentioned Pointed class. If it 
appears one day then your argument would translate to: We must acknowledge 
that 'pure' was always wrong in Applicative. Then we will have an APP 
(Applicative without 'pure' proposal) that again breaks lots of code 
without a way to code for multiple versions of base.